In modern industrial processes, there are numerous measure and control variables with serious coupling and complex relationships. Alternatively, distributed (multi-block) process monitoring methods can handle this problem for plant-wide process monitoring. However, the exchange of energy, materials and information exists among blocks in industrial processes. The distributed methods mentioned above always constructs a monitoring model in each block without communication from other blocks. Those modelling methods ignore the inherent connection between blocks and the monitoring models may be affected without the information from others. In order to overcome this limitation of the distributed methods, a distributed principal component analysis(PCA) process monitoring method based on the industrial process connection is proposed in this paper. Firstly, this method divides the plant-wide processes into blocks by prior knowledge. The topology matrix is proposed to describe the connection between blocks. In each iteration, the parameters of the monitoring model are updated according to the topology matrix using compressed information from other blocks and the final monitoring model is obtained. After substituting the tested data, the abnormal information carried by the residual blocks is taken into consideration in each blocks, and the final monitoring results are obtained. In the tennessee eastman process and the hydrocracking process, the proposed method is compared with the traditional multi-block PCA and fully connected distributed PCA method, which demonstrates the feasibility and effectiveness of the proposed method.